Having a store credit in $ means not U.K. I have checked and my 4K and later HD are HDMI only but my 3rd gen has Toslink so finding an older one will be cheap if your intent is purely as a Airplay endpoint.
I use an Apple TV 4K as a Roon endpoint and it works just fine via Airplay.
There’s also a 3rd party app on the App Store called TV:Remote that provides a nice display for Roon and provides some control and the ability to select what’s playing albeit in a less than ideal way.
I also have an older Apple TV with toslink. Beware that these don’t have access to the App Store so will work as a basic Roon endpoint but you won’t be able to load streaming apps or YouTube etc.
Why would the Apple TV “not sound so great?” The device isn’t doing ANY digital-to-analog conversion? It’s only taking an incoming Wi-Fi or Ethernet audio stream and outputting a digital stream that your DAC can handle.
If you’ve got a Gen.4 TV, then you only have HDMI out, but if you have a late G.3 one, you have TOSLINK optical out, and if you have an early G.3 one, you have both optical AND USB outputs.
If you have a modern asynchronous DAC, then the digital signal is reclocked (and all jitter removed).
I don’t understand how an Apple TV could have ANY effect on sound quality.
If you don’t need the actual video/TV part of it, why not get an old Airport Express? They were updated a couple years ago to support Airplay 2.0. I added one to my home network (disabled the Wi-Fi part) and it feeds the Toslink into a DAC. Sounds great to me.
Right off the bat, I have issues with “Apple Music Lossless Mess…”
This purports to be an investigation of Apple’s music transmission, but it’s being done with the “assistance” of a commercial competitor (Bluesound). What could possibly go wrong?
The results aren’t verified by any independent third party.
The examination sticks to the use of Apple sources and decoding. Some Bluesound gear was used, but there’s no explanation of WHY the Bluesound gear performed differently from the all-Apple chain. Were the differences caused by Apple gear? By Bluesound gear? Were there technical incompatibilities? What were they?
In short, even if the article is 100% accurate in its conclusions, there is not enough information given to understand why the conclusions given are correct (or not).
Compare the R&D budget of Bluesound to that of Apple. Who has more (and better) engineers? Does this disparity mean that Apple’s products are actually better than Bluesound’s? I’d think not. BUT the question remains, if Apple has the technical edge and the overwhelming budgetary edge, would they really introduce measurably inferior audio gear? They might. Audio may not be the priority to Apple that it is to Bluesound. But on the other hand, I’d expect Apple to be at least very, very close to Bluesound (and just about any other specialty manufacturer you’d care to name).
Having said that, could I be wrong? Of course I could. But without unbiased third-party evidence to the contrary, I’d think that Apple’s audio transmission protocols and hardware would be perfectly acceptable and competitive.
I have 3 endpoints that support both RAAT and AirPlay.
Playing the same 16/44 files from Roon, AirPlay sounds compromised to me.
It’s by no means bad or night and day different but the AirPlay streams have a glare or slight harshness that I find fatiguing over time. I primarily listen with headphones.
So while I’ll freely make reference to my username, I believe something is going on here.
This might explain it.
A quick question, please, regarding your comment that the Apple equipment sounds harsh compared to other endpoints? As mentioned elsewhere in this thread, it seems that Apple TV (don’t know about AirPort Express) automagically converts all incoming audio into 48kHz output.
Have you tried letting Roon do the conversion to 48kHz and then streaming it to the Apple device? This would (at least, theoretically) tax the Apple hardware significantly less and might eliminate the harshness that you are hearing.
Since Roon on a NUC should have clock cycles to spare, it would seem that converting at the NUC point in the signal chain would not stress the endpoint so much.